The eastern Beijing plain has been suffering severe subsidence for the last decades, mainly associated with the long-term excessive extraction of groundwater resource. Since the end of 2014, the annual water supply in Beijing plain has reached several hundred million cubic meters because of the South-to-North Water Diversion (SNWD) Project, which has reduced the groundwater exploitation and changed the status of land subsidence. In this work, we first obtain the current spatiotemporal variations of land subsidence in the eastern Beijing plain by using progressive small baseline subsets (SBAS) InSAR time series analysis method with Sentinel-1 SAR data acquired from July 2015 to December 2021. Then, we analyze the correlations between InSAR-derived subsidence and groundwater level change by applying the cross wavelet method. The results show that two major subsidence zones are successfully detected with the maximum deformation rate of −150 mm/yr and maximum cumulative deformation of −950 mm. Besides, the ground deformation at different stages from 2016 to 2021 reveal that the area and magnitude of major deformation significantly slow down, even in the regions with severe subsidence, especially in the year of 2017, which is about two years later than the start time of SNWD Project in Beijing. Further, we find the InSAR-derived subsidence lags groundwater level change with about 1–2-month lagging time, indicating that the dynamic variation of groundwater level fluctuation may be the main factor affecting the uneven subsidence in the severe subsiding zones. Last, differential subsidence rates are identified at both sides of geological faults, such as Nankou-Sunhe fault and Nanyuan-Tongxian fault, from the observed deformation map, which could be explained that the groundwater flow is blocked when a fault is encountered. These findings can provide significant information to reveal the deformation mechanisms of land subsidence, establish the hydrogeological models and assist decision-making, early warning and hazard relief in Beijing, China.

The rapid increase in point density and acquisition rate of UAV laser scanning (ULS) systems has shifted the primary bottleneck of LiDAR workflows from data acquisition to post-processing, particularly during direct georeferencing of massive LiDAR measurements. This study presents a systematic evaluation of parallel computing strategies for accelerating ULS direct georeferencing while preserving geodetic accuracy. Two georeferencing models are investigated: (1) a rigorous model that strictly follows the full geodetic transformation chain from sensor owned coordinates system (SOCS) to projected map coordinates, and (2) an approximate model that incorporates meridian convergence angle compensation and preprocessing of platform trajectories to reduce per-point computational complexity. For each model, a shared-memory multicore CPU implementation based on OpenMP and a heterogeneous GPU implementation based on CUDA are designed. Experiments were conducted on seven real-world ULS datasets, ranging from 2.9 × 10⁷ to 7.0 × 10⁸ points and covering diverse terrain types. Accuracy analysis shows that, in typical urban, plain, and industrial scenarios, the approximate model achieves millimeter-level mean errors and centimeter-level RMSEs relative to the rigorous model, satisfying the requirements of most engineering surveying applications. Performance evaluation demonstrates that parallelization yields substantial speedups: OpenMP-based method achieves 7–9 times acceleration, while GPU computing attains up to 24.6 times acceleration for the rigorous model and up to 16.7 times for the approximate model. The results highlight the complementary strengths of the two models and provide practical guidance for selecting accuracy-efficiency trade-offs in large-scale ULS production workflows.